A deploy takes a previously built artifact and stages it onto the target deployment platform. The target deployment platform is pluggable and can be Docker, Kubernetes, Nomad, EC2, and more.

"Stage" in this context means that the application should be ready to receive traffic, but is not yet open to public traffic. For example, the application should not be added to the load balancer, DNS should not be updated, etc. While it is staged, the deployment is accessible via the deployment URL or other internal means.

A deploy is triggered during a waypoint up or using the dedicated waypoint deploy command.


The deploy is configured using the deploy stanza within an app:

app "my-app" {
  deploy {
    use "kubernetes" {}

»Target Deployment Platforms

To deploy your app with Waypoint, you will need to select which platform to use. A deployment platform is a runtime for your app. It could be a container runtime capable of running a docker image or it could be a virtual machine platform. The deployment platform must be compatible with the app artifact output format from build.

You can currently use Waypoint to deploy your app to any of these platforms.

»How to Add a Deployment Platform to Waypoint

You can also create a plugin to extend Waypoint with your own deployment platform. If you would like to add a plugin for a platform currently not in Waypoint, please file a GitHub Issue for the project.

»Automatic Release

By default, waypoint deploy automatically performs a release.

This is because this is often the most expected behavior of a deploy since traditionally deploy and release were inseparable. We consider separating deploy and release steps a best practice so we encourage Waypoint users to disable this behavior as soon as they're comfortable.

This behavior can be disabled by using the -release=false flag with waypoint deploy:

$ waypoint deploy -release=false